Alright. I made a typo. The original code actually was:
OLD:
try {
Example_class *ec = new Example_class();
} catch (exception e) { //caught by value
cout << e.what();
}
In the new code, I just changed it to catch by reference.
NEW:
try {
Example_class *ec = new Example_class();
} catch(exception &e) {
cout << e.what();
}
I got a St9bad_alloc. Thats pretty much reasonable. My problem is
resolved.
On Fri, 12 Sep 2008, Anitha Boyapati wrote:
>
> Hi,
>
> Can somebody please throw light on "St9exception" being thrown during
> "new" operation ?
>
> I cannot put the source code here. Its simply huge. And it is not getting
> reproduced in smaller cases either. Here is a small description:
>
> //ALERT: NOT A TESTCASE!
>
> try {
> Example_class *ec = new Example_class()
> } catch(exception &e) {
> cout << e.what();
> }
>
> Somewhat euivalent code is given above. The output is St9exception.
> Example_class object gets constructed atleast 1 million times in my
> application successfully before it gives out an abort message from
> operator new.
>
> #0 0x007587a2 in _dl_sysinfo_int80 () from /lib/ld-linux.so.2
> #1 0x0079c7d5 in raise () from /lib/tls/libc.so.6
> #2 0x0079e149 in abort () from /lib/tls/libc.so.6
> #3 0x004f5557 in __cxa_call_unexpected () from /usr/lib/libstdc++.so.5
> #4 0x004f55a4 in std::terminate () from /usr/lib/libstdc++.so.5
> #5 0x004f5716 in __cxa_throw () from /usr/lib/libstdc++.so.5
> #6 0x004f5972 in operator new () from /usr/lib/libstdc++.so.5
>
> However at some point of time, when the application size really bulges out
> consuming 3.8GB of virtual memory, this exception is thrown and
> application is aborted [The application is a simulator, so it runs for
> quite a while]
>
> Now I would have been happy for a std::bad_alloc exception, but not with
> this St9exception. I can almost (ofcourse not 100% sure which is why I am
> here) say that there are no logical bugs or corrupted memories being
> accessed. Even if there are, a SIGSEGV should have been thrown instead of
> a SIGABRT.
>
> Thanks in advance!
>
>
>
--
Regards,
Anitha B
@S A N K H Y A
